SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

Hepatocarcinogenesis is a stepwise process. It involves several genetic and epigenetic alterations, e.g., loss of tumor suppressor gene expression (TP53, PTEN, RB) as well as activation of oncogenes (c-MYC, MET, BRAF, RAS). However, the role of RNA-binding proteins (RBPs), which regulate tumor suppressor and oncogene expression at the posttranscriptional level, are not well understood in hepatocellular carcinoma (HCC). Here we analyzed RBPs induced in human liver cancer, revealing 116 RBPs with a significant and more than 2-fold higher expression in HCC compared to normal liver tissue. We focused our subsequent analyses on the Insulin-like growth factor 2 messenger RNA (mRNA)-binding protein 1 (IGF2BP1) representing the most strongly up-regulated RBP in HCC in our cohort. Depletion of IGF2BP1 from multiple liver cancer cell lines inhibits proliferation and induces apoptosis in vitro. Accordingly, murine xenograft assays after stable depletion of IGF2BP1 reveal that tumor growth, but not tumor initiation, strongly depends on IGF2BP1 in vivo. At the molecular level, IGF2BP1 binds to and stabilizes the c-MYC and MKI67 mRNAs and increases c-Myc and Ki-67 protein expression, two potent regulators of cell proliferation and apoptosis. These substrates likely mediate the impact of IGF2BP1 in human liver cancer, but certainly additional target genes contribute to its function. Conclusion: The RNA-binding protein IGF2BP1 is an important protumorigenic factor in liver carcinogenesis. Hence, therapeutic targeting of IGF2BP1 may offer options for intervention in human HCC. (Hepatology 2014;59:1900–1911)

Abbreviations
c-MYC

v-myc myelocytomatosis viral oncogene homolog (avian)

FC

fold change

GFP

green fluorescent protein

HCC

hepatocellular carcinoma

IGF2BP

IGF2 mRNA-binding protein

iRFP

near-infrared fluorescent protein

NAFLD

nonalcoholic fatty liver disease

qRT-PCR

quantitative reverse-transcription, polymerase chain reaction

RBP

RNA-binding protein

shRNA

short hairpin RNA

siRNA

small interfering RNA

Hepatocellular carcinoma (HCC) is the fifth most common cancer and accounts for an estimated 695,900 deaths per year worldwide, representing the second most frequent cause of cancer-related death.[1-3] HCC accounts for 90% of all primary liver neoplasia and its incidence rate is increasing.[4] Long-term liver injuries caused by infection with hepatitis B or C virus, alcoholic liver disease, aflatoxin exposure, or inherited metabolic diseases contribute to the onset of HCC.[4] Hepatocarcinogenesis represents a multistep process in which tumor suppressor genes (TP53, PTEN, RB), oncogenes (c-MYC, RAS, MET, BRAF), developmental pathways (Wnt/β-catenin, Hedgehog), or growth factors and their receptors (IGF2, TGFB1, FGFR) are altered.[5, 6] Next to genetic aberrations, regulators of all layers of gene expression may control HCC-related genes and contribute to tumorigenesis.

RNA-binding proteins (RBPs) represent a large and diverse class of posttranscriptional regulators.[7] Through direct interaction, RBPs control the localization, stability, or translation of their target RNAs.[8] The insulin-like growth factor 2 mRNA-binding protein 1 (IGF2BP1), also known as IMP-1 or CRD-BP (coding region stability determinant-binding protein), belongs to a conserved family of RNA-binding, oncofetal proteins, which includes IGF2BP2 and IGF2BP3.[9] Several target RNAs of IGF2BP1 are known and often encode proteins that have exceptional roles in developmental processes and neoplastic transformation. For example, IGF2BP1 binds to the IGF2 mRNA and controls the translation of this growth factor.[10, 11] Binding of IGF2BP1 to ACTB messenger RNA (mRNA) regulates the spatiotemporal expression of ACTB in developing axons and dendrites.[12, 13] Interaction of IGF2BP1 with c-MYC and MDR1 mRNAs inhibits the endonucleolytic cleavage of these transcripts, prolonging mRNA half-life.[14-18] IGF2BP1 binds to the long noncoding RNA HULC and facilitates its degradation by way of interaction with CNOT1, a component of the deadenylation complex.[19] The expression of IGF2BP1 has been implicated in various cancers, e.g., breast, ovarian, brain, lung, pancreas, colon, and skin cancer as well as Hodgkin lymphoma and B cell lymphomas.[9] Only recently, the family member IGF2BP2 was reported to be up-regulated in liver cancer,[20] but neither the regulation nor the specific functional role of IGF2BP1 in HCC has been studied.

Here we show that IGF2BP1 is the most strongly up-regulated RNA-binding protein in human HCC compared to normal liver. Its depletion in several liver cancer cell lines inhibits proliferation and induces apoptosis, partially by way of down-regulating c-MYC mRNA and protein levels. As a novel target for IGF2BP1, we identify the proliferation marker protein MKI67 (Ki-67). Importantly, a stable reduction of IGF2BP1 in HepG2 cells impairs tumor growth in vivo in a murine xenograft model. Thus, our study identifies IGF2BP1 as a novel potential target in HCC treatment.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information
RNA Isolation, cDNA and Quantitative Reverse-Transcription Polymerase Chain Reaction (qRT-PCR) Analysis

Trizol lysis and RNA isolation was done as described previously.[21] For cDNA synthesis, equal amounts of RNA were transcribed and random primers (Life Technologies) were used for reverse transcription as described previously.[22] Gene expression was measured on an ABI StepOnePlus using SYBRGreen (Life Technologies). The housekeeping gene GAPDH was used as reference gene in all qRT-PCR analyses. All qRT-PCR primers are shown in Supporting Table 2.

RNA Interference

RNA interference (RNAi) was essentially performed as described previously with minor modifications.[19] In brief, for siRNA-mediated gene knockdown, 1 × 105 (HLE, HLF, Huh6) or 2 × 105 cells (HepG2, Huh7, Hep3B) were reverse transfected with 5 μL small interfering RNA (siRNA) (20 μM) and 5 μL RNAiMAX (HepG2, Huh7, Hep3B) or 10 μL siRNA (20 μM) and 10 μL RNAiMax (Life Technologies) in a 6-well plate according to the manufacturer's recommendations. Cells were harvested 48 hours or 72 hours after transfection for subsequent gene expression analysis (protein or RNA).

Sequences of the individual siRNAs can be found in Supporting Table 3, shRNA sequences for lentiviral transduction in Supporting Table 4. As nontargeting siRNA control, siAllStars Negative Control siRNA from Qiagen (Hilden, Germany) was used.

Proliferation Assays

Proliferation was analyzed with the bromodeoxyuridine (BrdU) Cell Proliferation ELISA kit (Roche, Basel, Switzerland) as described previously[23] or by monitoring cell numbers by counting at different timepoints (TC-20 Cell Counter, Bio-Rad).

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information
IGF2BP1 Is the Most Strongly Up-Regulated RNA-Binding Protein in HCC

Unbiased microarray analysis of 60 human HCC (Supporting Table 1) and seven normal liver samples was performed previously using the Agilent SurePrint G3 Human Gene Expression array.[24] Here we focused our analysis on RBPs in human HCCs as defined by the Gene Ontology terms “RNA binding” (GO:0003723) and “mRNA binding” (GO:0003729). This yielded a list of 116 significantly up-regulated (fold change [FC] ≥ 2.0; P ≤ 0.05) RBPs and identified IGF2BP1 as the most highly up-regulated RBP in HCC (FC = 6.8, P = 7.8 × 10−3, t test) (Fig. 1A; Supporting Table 5). We confirmed the overexpression of IGF2BP1 mRNA in HCC with qRT-PCR (Fig. 1B). Interestingly, IGF2BP1 expression positively correlated with tumor size (P = 0.038; correlation coefficient [Spearman-Rho]: 0.276). We also tested for correlation with other clinical parameters (etiology, staging, grading, vascular invasion, age, and sex), but only detected weak but significant correlations with advanced stages and grading of poor differentiation (Supporting Fig. 1, Supporting Table 1).

image

Figure 1. Differential expression of IGF2BP1 in HCC and normal liver patient samples. (A) Microarray analysis of 60 HCC and 7 normal liver samples showing significantly increased IGF2BP1 mRNA expression in HCC. The horizontal line represents the mean of nontumor samples. (B) Validation of differential IGF2BP1 expression using qRT-PCR. GAPDH was used as reference gene. The boxplot shows a significantly higher expression of IGF2BP1 mRNA in tumor samples (P = 0.03). (C) Confirmation of IGF2BP1 up-regulation at the protein level. Tissue microarray staining was performed using a custom-made anti-IGF2BP1 antibody (clone 6A9). A strong cytoplasmic staining in liver cancer samples is detected (lower panel). (D) Analysis of protein expression using the immunoreactive score reveals higher IGF2BP1 protein levels in HCCs.

Download figure to PowerPoint

In addition to our microarray data, we confirmed IGF2BP1 induction in the Oncomine database (www.oncomine.org) (Supporting Fig. 2A,B). Analysis of the Chen et al. liver dataset[25] confirmed a higher expression of IGF2BP1 in HCC compared to normal liver tissue and precursor lesions. The comparison of ∼900 human cancer cell lines of different origin[26] revealed the highest expression level of IGF2BP1 in liver cancer cells, implicating a specific importance of IGF2BP1 in this entity (Supporting Fig. 2C).

To confirm the induction of IGF2BP1 at the protein level, we performed tissue microarray analyses and stained 101 HCCs and 11 nontumorous liver samples (Fig. 1C,D). We observed a stronger cytoplasmic staining for IGF2BP1 in the HCC samples compared to the nontumorous sections recapitulating the up-regulation of IGF2BP1 mRNA also at the protein level. The IGF2BP1 antibody used was custom-made and specifically recognized human IGF2BP1 in western blotting (Supporting Fig. 3).[9]

Depletion of IGF2BP1 Inhibits Proliferation and Induces Apoptosis

IGF2BP1 was highly expressed in a panel of six human liver cancer cell lines, both at the mRNA and protein level (Fig. 2A). Only minor expression differences were detected. Next we analyzed the cellular loss-of-function phenotype by way of siRNA-mediated depletion of IGF2BP1 with two independent siRNAs. First, the proliferation of the six cell lines was measured using a BrdU incorporation assay. Loss of IGF2BP1 significantly decreased proliferation in all cell lines (Fig. 2B). The effect was most pronounced in HepG2 cells (∼10-fold), while Hep3B cells showed only a 2-fold reduction in proliferation. The reduced proliferation rate was accompanied by a significant reduction of the proliferation markers CCND1 and PCNA in all six cell lines as determined by qRT-PCR (Fig. 2C; Supporting Fig. 4A). Moreover, depletion of IGF2BP1 efficiently induced apoptosis in all six cell lines as indicated by enhanced poly (ADP-ribose) polymerase (PARP) cleavage (Fig. 2D; Supporting Fig. 4B).

image

Figure 2. Reduced proliferation and enhanced apoptosis after IGF2BP1 depletion. (A) IGF2BP1 mRNA and protein expression were analyzed in six human liver cancer cell lines. GAPDH was used as reference gene in qRT-PCR analysis as well as loading control for western blot. (B) A BrdU incorporation assay was used to analyze cellular proliferation 72 hours after IGF2BP1 knockdown with either of two siRNAs. The proliferation rate was normalized to siControl for each cell line. (C) Expression analysis of proliferation markers CCND1 and PCNA after 48 hours or 72 hours of IGF2BP1 depletion. The expression was normalized to siControl and GAPDH was used as reference gene. (D) Analysis of apoptosis induction 72 hours after IGF2BP1 knockdown. Total protein lysates were analyzed for enhanced PARP cleavage as a marker for apoptosis by western blotting. Efficient knockdown of IGF2BP1 was accomplished in all cell lines. Representative western blots are shown. All experiments were done in biological replicates (n=3). Given is the mean and the respective standard error of the mean (±SEM).

Download figure to PowerPoint

Knockdown of IGF2BP1 Impairs Tumor Growth In Vivo

As knockdown of IGF2BP1 strongly reduced cancer cell proliferation and survival in vitro, we tested whether depletion of IGF2BP1 also prevented tumor growth in vivo. HepG2 cells were transduced with lentiviral vectors encoding IGF2BP1-directed shRNA and an iRFP expression cassette to trace transduced cells. Efficient IGF2BP1 knockdown was confirmed by western blot (Fig. 3A, upper panel). Similar to transient IGF2BP1 depletion, shRNA-mediated knockdown of IGF2BP1 correlated with reduced cell proliferation (Fig. 3A, lower panel). Subcutaneous injection of 4-8 × 105 HepG2 cells transduced with the control shRNA vector (sh-Control) led to the formation of macroscopic tumors in 10 out of 16 animals (Fig. 3B). In contrast, only 3 out of 16 animals showed macroscopic tumors when injected with HepG2 cells depleted of IGF2BP1 (sh-IGF2BP1). This was analyzed in further detail by near-infrared (NIR) in vivo imaging of iRFP-labeled tumor cells. Surprisingly, the in vivo imaging revealed that the total number of micro- and macrotumors was the same in both populations with 13 out of 16 tumors (Fig. 3B). However, in the IGF2BP1 knockdown cohort, the number of microscopic and thus nonpalpable tumors was significantly increased. In agreement, Kaplan-Meier analyses confirmed a significantly delayed formation of macroscopic tumors (Fig. 3C). This was further supported by the quantitative assessment and significant decrease of endpoint tumor area and integrated fluorescence intensities (FI) as determined by NIR-imaging (Fig. 3D-F). These findings provide the first evidence that IGF2BP1 depletion impairs tumor growth in vivo and support in vitro observations indicating that the protein promotes tumor cell proliferation and survival.

image

Figure 3. IGF2BP1 is important for tumor growth in vivo. IGF2BP1 depletion impairs the growth of heterologous xenograft tumors. (A) HepG2 cells were transduced with the indicated multiplicity of infection (MOI) of control or IGF2BP1-directed shRNA encoding lentiviral vectors for 72 hours. IGF2BP1 depletion was monitored by western blotting using the indicated antibodies (upper panel). VCL served as a loading control. Cellular proliferation relative to controls transduced with control shRNA (sh-Control) was determined by cell counting (lower panel). Statistical significance was determined by Levene's F- and Student t test. (B-F) Transduced HepG2 were harvested 48 hours after transduction and suspended in serum-free media containing Matrigel before subcutaneous injection (4-8 × 105 cells per animal). For each cohort, control knockdown (sh-Control) and IGF2BP1 depletion, 16 animals were injected. Tumor growth was monitored at 5-day intervals. The number of macroscopic palpable tumors as well as microtumors was determined 115 days (endpoint) after injection by NIR-imaging (B). Time resolved macroscopic tumor burden is depicted by a Kaplan-Meier plot (C). The area (D) and integrated fluorescence intensity (E) of macro- and microtumors identified in both cohorts is depicted by boxplots. Images of three macroscopic tumors acquired by brightfield (DIA) or NIR-imaging (iRFP) are shown in (F). Bar = 10 mm; LUT, fluorescence intensity scale (white = highest / black = lowest). Statistical significance was determined by logrank test (C) or Mann-Whitney U tests (E,F).

Download figure to PowerPoint

IGF2BP1 Regulates c-MYC Expression at the Posttranscriptional Level

To gain further insights into the molecular mechanisms underlying IGF2BP1's cellular functions, we hypothesized that IGF2BP1 may posttranscriptionally regulate the expression of oncogenes or tumor suppressors that are known to have pivotal roles in hepatocarcinogenesis. Among the known IGF2BP1 target genes, e.g., ACTB,[13] IGF2,[10, 11] MYC,[14, 15] or MDR1,[17] MYC represented the most interesting target due to its established role as regulator of proliferation and apoptosis in diverse human cancers including HCC.[27-30] Notably, previous studies revealed that IGF2BP1 promoted the proliferation of ovarian cancer cells by enhancing the expression of c-MYC mRNA and protein. This was correlated with a severe up-regulation of IGF2BP1 expression in serous ovarian carcinomas.[31] Thus, we analyzed the expression of c-MYC after IGF2BP1 depletion in HepG2 cells. The mRNA and protein levels of c-MYC were significantly down-regulated upon IGF2BP1 knockdown (Fig. 4A,B). This was also observed in Huh6 cells (Supporting Fig. 5A,B). To confirm that IGF2BP1-dependent enhancement of c-MYC expression was correlated with a direct association of IGF2BP1 with the c-MYC mRNA, we performed RNA immunoprecipitation (RIP) experiments. FLAG-tagged IGF2BP1 or GFP (green fluorescent protein; negative control) were transiently overexpressed in HepG2 cells and immunoprecipitated with an anti-FLAG antibody (Fig. 4C, left panel). After isolation of the copurifying RNA, the enrichment of selected transcripts was determined by way of qRT-PCR. This confirmed selective enrichment of the c-MYC mRNA as well as other previously identified IGF2BP1 target transcripts, i.e., the IGF2 mRNA and the long noncoding RNA HULC (Fig. 4C, right panel). No enrichment of c-MYC mRNA was seen in GFP controls. The highly abundant 5.8S rRNA or lysine-tRNA served as negative controls and were not enriched in any of the purifications. Thus, we confirmed selective association of IGF2BP1 with the c-MYC mRNA in human liver cancer cells.

image

Figure 4. Impact of IGF2BP1 depletion on c-MYC expression and mRNA stability. (A) Analysis of c-MYC mRNA 48 hours after IGF2BP1 depletion in HepG2 cells. GAPDH was used as reference gene in qRT-PCR. Depicted is the remaining c-MYC expression relative to its level in cells transfected with control siRNA (mean of at least three independent experiments ±SEM). (B) A representative western blot shows the reduction of c-Myc protein after IGF2BP1 depletion. (C) Representative western blot analysis after FLAG-immunoprecipitation. HepG2 cells were transiently transfected for 72 hours with FLAG-tagged GFP (negative control) or IGF2BP1, respectively (left panel). Analysis of copurified RNA and respective enrichment as determined by qRT-PCR after anti-FLAG immunoprecipitation showing the specific binding of c-MYC mRNA to IGF2BP1 (right panel). All immunoprecipitation experiments were done in biological replicates (n = 3). Given is the mean and the respective standard error of the mean (±SEM). HC, heavy chain. (D) c-MYC mRNA stability analysis in HepG2 cells after actinomycin D (ActD) treatment. Cells were transfected with siRNAs against IGF2BP1 or a control siRNA and 48 hours later a time course for RNA stability was started by adding the transcription inhibitor. Cells were harvested at the indicated timepoints. Expression levels were normalized to “0 h” and GAPDH was used as reference gene. Shown is the mean of at least three independent experiments (±SEM).

Download figure to PowerPoint

Since IGF2BP1 could stabilize, destabilize, or control the translation of its target RNAs, we tested whether its knockdown affected c-MYC mRNA stability in liver cancer cells as previously shown in other tumor entities.[18] Upon siRNA-directed IGF2BP1 depletion, HepG2 cells were treated with Actinomycin D to inhibit transcription. Indeed, we observed an enhanced decay and a significantly reduced half-life (siControl: ∼43 minutes versus siIGF2BP1: ∼32 minutes) of c-MYC mRNA (Fig. 4D). Hence, IGF2BP1 directly associated with the c-MYC mRNA, stabilized this target transcript, and gave rise to an enhanced c-MYC expression.

Since the c-MYC oncogene is a key player associated with malignant transformation by regulating cell cycle progression, differentiation, cell growth, apoptosis, and other tumor traits,[28] we tested whether c-MYC down-regulation could contribute to the observed cellular phenotypes associated with IGF2BP1 depletion. Two different siRNAs efficiently reduced c-MYC at the mRNA and protein level (Supporting Fig. 6A,B). At the cellular level, c-MYC knockdown reduced proliferation of HepG2 cells, which correlated with the knockdown efficiency (Supporting Fig. 6C). In addition, the depletion of c-MYC slightly induced apoptosis in a dose-dependent manner in HepG2 cells, as indicated by enhanced PARP cleavage (Supporting Fig. 6B). Knockdown of c-MYC with the second siRNA was apparently not sufficient to induce apoptosis, but consistently affected cell proliferation (Supporting Fig. 6C). A similar decrease in proliferation upon c-MYC knockdown was obtained in Huh6 cells (Supporting Fig. 6D-F).

To determine whether c-MYC could, at least partially, rescue the cell growth phenotype observed upon IGF2BP1 depletion, we overexpressed c-MYC at moderate levels in IGF2BP1-depleted cells (Supporting Fig. 5C). Consistent with our previous findings, IGF2BP1 knockdown strongly impaired cell growth. The latter was substantially restored by the overexpression of Flag-tagged c-MYC protein, suggesting that IGF2BP1 promotes tumor cell growth and survival partially by way of c-MYC (Supporting Fig. 5D). Notably, IGF2BP1 was modestly up-regulated by the overexpression of c-MYC even in the presence of IGF2BP1 targeting siRNAs. This suggested that c-MYC could induce IGF2BP1 expression. Depletion of c-MYC in HepG2 and Huh6 cells indeed reduced IGF2BP1 mRNA and protein expression and its overexpression strengthened IGF2BP1 promoter activity (Supporting Fig. 7), supporting previous studies suggesting that c-MYC promotes IGF2BP1 mRNA synthesis.[32] This might point towards a positive feedback loop and further underscores the functional interplay between these two factors in liver cancer.

Finally, we tested whether IGF2BP1-directed enhancement of c-MYC expression was also detectable in human HCCs. We correlated IGF2BP1 and c-MYC mRNA expression in HCC patient samples. Consistent with IGF2BP1-dependent control of c-MYC expression in vitro, we observed a modest, but significant, positive correlation between the expression of IGF2BP1 and c-MYC mRNA in our microarray dataset (P = 0.029; correlation coefficient [Spearman-Rho]: 0.282) (Supporting Fig. 8A). However, this analysis revealed two distinct groups of HCC with different interdependencies between IGF2BP1 and c-MYC (Supporting Fig. 8B,C): A subset of HCCs with well detectable IGF2BP1 expression (group A, red dots) showed a strong and significant correlation (P = 0.013; correlation coefficient 0.558). In contrast, group B HCCs (gray dots) with marginal expression of IGF2BP1 did not show a significant correlation (P = 0.446; correlation coefficient 0.122). Hence, in HCCs with relevant expression of IGF2BP1, its expression was positively correlated with c-MYC expression. We compared the gene expression profiles between the two groups (A, B) and analyzed pathway activation using the DAVID v. 6.7 analysis tool (http://david.abcc.ncifcrf.gov/)[33] for genes that showed a significant (P < 0.05) and at least a 1.5-fold increase in group A versus group B. Pathways overrepresented in group A were most prominently linked to the cell cycle (Supporting Fig. 8D). These processes might thus be distinct between the two groups of HCCs with high versus marginal IGF2BP1 expression and could be causally linked to IGF2BP1 and c-MYC expression.

Ki-67 Is a Novel Target for IGF2BP1 in Liver Cancer

While c-MYC is undoubtedly an important target of IGF2BP1 with a strong link to cancer, a direct comparison of the proliferative and apoptotic effects after IGF2BP1 and c-MYC silencing (Fig. 2B; Supporting Fig. 6) revealed a stronger impact of IGF2BP1, which implicates additional targets in its function in liver cancer cells. Hence, we aimed to explore further targets of IGF2BP1 in HCC. We identified mRNAs positively correlated with IGF2BP1 in our HCC microarrays and compared these to the mRNAs interacting with IGF2BP1 in public datasets.[34] Thereby, we identified the proliferation marker MKI67 (Ki-67)[35] as a novel client mRNA of IGF2BP1. Ki-67 mRNA was strongly up-regulated in HCC compared to normal liver (Fig. 5A) and positively correlated with IGF2BP1 expression (Fig. 5B). Additionally, we found a positive, but not statistically significant, correlation between IGF2BP1 protein expression and the number nuclei positive for Ki-67 protein in the HCC TMA (P = 0.09, correlation coefficient [Spearman-Rho]: 0.345). Using RIP-qPCR, the interaction of Ki-67 mRNA with IGF2BP1 protein was validated (Fig. 5C). Upon IGF2BP1 silencing, Ki-67 mRNA as well as protein was reduced (Fig. 5D-F). At the mechanistic level, depletion of IGF2BP1 significantly decreased the half-life of Ki-67 mRNA (Fig. 5G).

image

Figure 5. Impact of IGF2BP1 depletion on MKI67 expression and mRNA stability. (A) MKI67 mRNA levels (represented by three different probes) were significantly increased in HCC tissue compared to normal livers as detected by microarray analysis of 60 HCC and 7 normal liver samples. (B) Positive and strong correlation of IGF2BP1 and MKI67 mRNA in 60 HCC samples. (C) Analysis of copurified RNA and respective enrichment as determined by qRT-PCR after anti-FLAG immunoprecipitation showing the specific binding of MKI67 mRNA to IGF2BP1. (D) MKI67 mRNA levels were significantly reduced in HepG2 and Huh7 cells with two independent siRNAs as analyzed by qRT-PCR. (E) The reduced MKI67 protein was analyzed by western blotting. Tubulin was used as reference gene. All experiments were done in three biological replicates and representative blots are shown. (F) IGF2BP1 (green) and MKI67 (red) protein expression were determined by immunofluorescence in Huh7 cells after knockdown of IGF2BP1 (blue: DAPI; scale bar = 25 μm). (G) MKI67 mRNA stability analysis in HepG2 cells after alpha-amanitin treatment. Cells were transfected with siRNAs against IGF2BP1 or a control siRNA and 48 hours later a time course for RNA stability was started by adding the transcription inhibitor. Cells were harvested at the indicated timepoints. Expression levels were normalized to “0 h” and GAPDH was used as reference mRNA. Shown is the mean of at least three independent experiments (±SEM). *P < 0.05; **P < 0.01; ***P < 0.001.

Download figure to PowerPoint

Taken together, we propose the following model of IGF2BP1 function in human HCC (Supporting Fig. 9): The RNA-binding protein interacts with several protein-coding mRNAs, including c-MYC and MKI67, or long noncoding RNAs and regulates their fate, i.e., degradation, translation, and/or localization. Interaction with c-MYC and MKI67 mRNAs leads to a stabilization and increased expression in liver cancer. Altogether, these regulatory circuits that are controlled by IGF2BP1 act in concert to allow tumor cell proliferation, tumor growth, and prevent apoptosis in liver cancer.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

Large-scale genomic approaches have identified genetic abnormalities in putative “drivers” of transformation in liver cancer.[36-39] HCC appears as a multifactorial disease emerging from a stepwise process including loss of tumor suppressors on chromosome 1, 3, and 8p, and gain of oncogenes such as c-MYC on 8q.[5, 40, 41] Nonetheless, the molecular mechanisms underlying neoplastic transformation are incompletely understood. Oncogenic drivers without chromosomal alteration or mutation are missed in these analyses, although their expression may be deregulated. Moreover, cellular processes are dynamically regulated also posttranscriptionally and posttranslationally. MicroRNAs are well established as posttranscriptional regulators in HCC.[42] However, surprisingly little is known about RNA-binding proteins and their function in liver carcinogenesis.

RNA-binding proteins can influence gene expression by their direct interaction with specific sequence motifs or structural elements present in the coding or 5′- or 3′-untranslated region of their target RNA[7, 8] controlling RNA localization, stability, or translation. So far, only a small number of RBPs have been studied in human liver diseases, e.g., HuR,[43] CUGBP1,[44] and Lin28.[45]

In this study we determined the role of IGF2BP1 in liver cancer. In our microarray profiling, IGF2BP1 is the most highly up-regulated RBP in human HCCs (Fig. 1). The precise mechanism of regulation remains elusive. IGF2BP1 gene amplifications have been found in human breast cancer.[46] However, no evidence of IGF2BP1 genomic alterations (17q21.32) in liver cancer is found in public databases (COSMIC, CONAN, Oncomine), suggesting a transcriptional induction. Notably, the IGF2BP2 family member is not strongly induced in our patient cohort in contrast to recent observations.[20] This finding is consistent with reports that IGF2BP2 rather plays a role in metabolic control than in carcinogenesis.[9, 47] Also, while IGF2BP2 silencing has been published to reduce ERK phosphorylation as its main mechanism of action,[20] we do not observe this effect for IGF2BP1 silencing (Gutschner, Hämmerle, Diederichs, unpublished). This further underlines the functional distinctness of these family members.

IGF2BP1 is a known oncofetal protein linked to several human cancers: Its expression is induced in human malignant melanomas or colorectal carcinomas with activated WNT/β-catenin/TCF signaling.[16, 48] High IGF2BP1 expression is a poor prognostic marker in ovarian and lung cancer.[31, 49, 50] Here we show that IGF2BP1 expression in HCC patients is associated with enhanced tumor growth and identify IGF2BP1 as a critical regulator of liver cancer cell proliferation and survival (Fig. 2). Our xenograft assay further supports the notion that IGF2BP1 is important for liver cancer growth in vivo (Fig. 3). The correlation of IGF2BP1 expression with tumor size and the formation of microtumors in the xenograft mouse model might indicate that IGF2BP1 is important for liver cancer progression and tumor growth rather than initiation.

IGF2BP1 may contribute to liver carcinogenesis in part by regulating the expression of the oncogene c-MYC. IGF2BP1 directly interacts with the c-MYC mRNA, stabilizes the transcript, and leads to higher c-MYC protein expression (Fig. 4). These findings corroborate previous findings from other cell types in liver cancer.[18, 31]

c-MYC acts as a transcription factor and controls the expression of pro-oncogenic factors that drive cell proliferation.[28] Interestingly, c-MYC also seems to moderately induce IGF2BP1 expression and promoter activity in HepG2 and Huh6 cells (Supporting Fig. 7), recapitulating findings in HeLa and HEK293T cells[32] in a liver cancer model. Chromatin immunoprecipitation data deposited in the UCSC genome browser (www.genome.ucsc.edu) provide evidence for direct binding of the c-Myc/Max dimer to the promoter region of IGF2BP1 in HepG2 cells, further supporting the idea of a self-amplifying loop. This feedforward loop may contribute to the maintenance of high c-MYC levels in liver cancer cells.

Depletion of c-MYC leads to decreased cell proliferation and induced apoptosis (Supporting Fig. 6). While this might seem unexpected at first, given that c-MYC overexpression is also known to induce apoptosis, our finding matches previous observations in liver cancer cells.[29] Loss of c-MYC or IGF2BP1 give rise to similar phenotypes, but our data also suggest that c-MYC is not the sole mediator of IGF2BP1 function. Concordantly, IGF2BP1 depletion reduces c-MYC expression in HepG2, Huh6, and Huh7 cells, but not in other liver cancer cell lines tested (data not shown), although these are highly responsive to IGF2BP1 knockdown. This could be explained by the heterogeneity and mutational diversity of the different cell lines or additional factors that control c-MYC expression levels and render it largely insensitive towards changes in IGF2BP1 expression. Thus, besides its striking effect on c-MYC expression in a subset of liver cancer cells, IGF2BP1 likely achieves its full oncogenic potential by the pleiotropic regulation of multiple target genes. Here we identify one additional substrate mRNA for IGF2BP1: MKI67 (Ki-67). The Ki-67 mRNA is induced in HCC and positively correlated with IGF2BP1, it interacts with IGF2BP1, its expression is reduced, and its mRNA destabilized upon IGF2BP1 knockdown. Ki-67 has been linked to proliferation, one of the hallmarks of cancer, providing a further important link between IGF2BP1 and its oncogenic phenotypes.

Future studies will elucidate the full RNA-IGF2BP1 network that promotes HCC development.

In summary, our study unravels a striking example that RNA-binding proteins represent important functional regulators of liver carcinogenesis by defining IGF2BP1 as an important player in human HCC with a broad impact on proliferation, survival, and tumor growth. Hence, inhibition of IGF2BP1 expression or blocking the interaction with its target RNAs may represent a promising and innovative approach for clinical intervention.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

The authors thank Dr. Georg Stoecklin for helpful discussions and Dr. Vladimir Spiegelman for IGF2BP1 promoter luciferase constructs. We thank Karin Rebholz for excellent technical assistance with immunohistochemical stainings and Dr. Bernhard Hiebl for mouse injections. This project was supported by the tissue bank of the National Center for Tumor Diseases (NCT) of Heidelberg.

Author Contributions

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

T.G. and M.H. conceived the study, designed and executed the in vitro experiments, analyzed and interpreted the data, and wrote the article. N.P. designed, executed, and analyzed in vivo xenograft experiments and helped write the article. N.B., E.F., H.U., A.H., M.G., N.H., B.M. performed experiments and analyzed data. B.S. and R.G. performed microarray expression analysis. T.L., K.B., and P.S. provided tissue samples and helped with data interpretation. S.H. and S.D. contributed to the design of experimentation and data interpretation and wrote the article.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article at the publisher's website.

FilenameFormatSizeDescription
hep26997-sup-0001-suppinfo.pdf514KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.